Saturday, August 23, 2008

CDN discussion on WebHostingTalk

CDN - any demand at all?

Guys,
I am wondering - how many of you guys are considering buying into a CDN service? Do you ever find yourself (or your clients) asking for it?

If so, in what specific situations would you need a CDN? Did you already buy the service - if so, what made you choose your provider, and if you did not buy the service yet, why not?


D

__________________
Ditlev Bredahl, CEO
www.UK2group.com

---

Quote:
If so, in what specific situations would you need a CDN? Did you already buy the service - if so, what made you choose your provider, and if you did not buy the service yet, why not?
Great post eming!

My experience in selling CDN has been well varied across a number of situations. Considerations to include are number of files to be cached, average file size, and number of downloads per month. Anything delivered via HTTP or Streaming can utilize a CDN. Specific situations vary from full time whole site caching to one time events and everything in between.

As a general rule if you are pushing less than 1TB of traffic a month you do not need a CDN. However if you have events every few months that spike above this or you are planning on launching a new marketing program a CDN is a good thing to consider.

Youtube started with a few CDN's well before they took off, but it was their use of CDN that kept their Quality of Experience high enough to keep users coming back. You can have the coolest website in the world but if it starts lagging or crashing because of a sudden surge in use then everyone is gonna move onto the next website of the week.

Looking at your existing situation and then putting together a 3 month, 6 month, and 1 year growth plan is essential. If you can show a CDN that you are serious and have plans to grow you can negotiate more aggresive pricing.

Right now the industry is in a flux. If you read Akamai's earnings call they are blaming the slow in the US economy on a lag in their growth, however they are still showing 19% quarterly growth. Limelight is still burning cash too fast to make a profit but they are serving up the Olympics to the US and its pretty impressive stuff. Level 3 is still building their network out but have a strong advantage in that they own the network so they can get really aggressive on pricing.

__________________
------------------------------
Random bits mixed with sand
desertstandard

---

I actually work for the largest CDN (Akamai) and I can tell you that CDNs are a requirement for a lot of today's highly interactive web and business applications; I would go so far as to say the internet as we know it.

Streaming HD video to Australia or running your online business portal globally would be impossible without today's CDNs. There simply is not enough effective bandwidth on the internet. Serving content from inside the last mile on the ISP's local network this coupled with advanced mapping and routing algorithms is by far the fastest and most efficient way to deliver content.

Which is another thing that should be considered. All CDNs are not created equal. Limelight and all of the other budget CDNs do not have the reach like Akamai does. You also get what you pay for: some DCs / CDN providers claim 3, 4 or even five 9's of uptime. That is all for beginners; we measure uptime in 0's. (All 0's = 100.00% uptime.) Not to mention our ability to move large amounts of data seamlessly.

Limelight is streaming the Olympics to the US, however Akamai is streaming it to the rest of the world. Limelight does not have the reach to do that. If right now there were to be some other major media event on their network it would most likely crumble. To see a 300+ gb/sec spike in traffic on our network is not uncommon but it makes no difference to our daily operation. On to the next topic at hand...

Polling people here about their customers' use of CDNs is useless. You're not going to find the decision makers of Yahoo, Amazon, MySpace or Disney trawling the forums here looking for budget CDNs. They're going to be going right for the big dogs. And as far as things go here on this forum; we're all small players. No one here is doing 200+ million dollar deals with bandwidth usage statistics that DCs like Softlayer and FDC could only dream of hitting. (IE the previously aforementioned 300gbps spike on top of our normal daily traffic)

Just my perhaps bias and slightly cynical $0.02

__________________
Echoservers.com
Premium Dedicated, VPS and shared hosting.
Never Oversold.

---

CDN is growing big time to say otherwise is to miss what is happening in the marketplace. Annual growth for the industry is around 30%.

Its not just the Yahoo's and MSN's that are using CDN.

I am not a regular user of this forum so I may be out of touch with who is reading these posts. Still it is a very good conversation to have as CDN is becoming more and more prevalent. If you are building a website/platform that you want to grow and make money with then CDN will enter the picture eventually.

That said if you are on a shared hosting account and are looking at CDN you have skipped a step, and that step is a dedicated server. I have many clients that came to me about CDN services while they were still on a shared server. The first step was to get them to one of my Partners and onto a Dedicated solution, then we rolled them into CDN.

__________________
------------------------------
Random bits mixed with sand
desertstandard

----

Softlayers CDN which is really just a resold service from InterNap is really good.

Saves me a lot of hassle

__________________
--
Damien Jorgensen
www.damienjorgensen.co.uk

---

Quote:
Originally Posted by lockbull View Post
I guess I'm in the minority here, but I think the bloom is kind of off the CDN rose. Akamai and Limelight reported disappointing results, Internap is having major integration and revenue problems with it's CDN, and analysts are predicting there are way too many struggling venture backed CDNs compared to the anticipated market revenue. Plus there are still some unsettled intellectual property issues in this space. My guess is that if you wait around say 12-24 months you'll be able to pick the carcasses of several dead CDNs at an attractive price.
making a 20% growth quarter over quarter is anything but a failing business. missing the Street's expectations is nothing new. i agree that the CDN market is being flooded at the moment and that a shakeout is inevitable but this is how business works.

the fact remains that content growth is exploding online and without a caching service to help things along users will have to get used to slow/spotty/latency riddled delivery or providers will have to build out more and more servers to keep up with the demand. users will not accept this and unless you are a tech company who wants to build a network then the providers are not interested in building and maintaining such a network. it also doesnt make sense to build out something that can handle your maximum traffic but is underutilized the majority of the time. the flexibility of being able to handle a mention on Operah is not an everyday need, but when it happens you want to be able to handle it with no downtime. this is what a cdn gives you.

No comments: